In this project, the next 60 days of Rossmann's (Germany) average daily sales have been predicted using
time series analysis and hybrid boosting with a Root Mean Square Percentage Error: 0.1774.
Forecasting by using the SARIMAX approach is underway.
I created a dashboard of Rossmann stores sales data, and analysed the sales by;
- State
- Store
- Store Type
- Assortment Level
As a fellow anime fan, I wanted to make an anime recommendation system as a portfolio project.
To achieve this, I created a web scraper by using Jikan (時間) REST API to scrap 2334 TV anime productions data between 2010-2022 and 86269 individual users' data from the most used online anime and manga catalogue myanimelist[DOT]net.
In this project, I cleaned the given housing data of Nashville by using MySQL queries.
The project is inspired from a Microsoft SQL tutorial, I rewrote the query tasks in the tutorial in MySQL queries.
In this project, I applied multiple machine learning models on oxide content and reflexion index values of various glasses and classify them through 7 types of glasses. The most accurate model appeared as Extra Trees Classifier with 0.82 accuracy score.
I used Rossmann store sales data between 30.09.2015-12.12.2015 to create a simple, one page dashboard to analysed the periodical sales by;
- State
- Store Type
- Assortment Level
- ChatGPT Prompt in Action: An End-to-End Generic DS Project
- Anime Recommendation System Through Hybrid Clustering
- Deep-Learning project using Tensorflow/PyTorch